...
- Add int8 scalar quantization to the HNSW vector format. This optionally allows for more compact lossy storage lossy storage for the vectors, requiring about 75% memory for fast HNSW search. (Ben Trent)
- HNSW graph now can be merged with multiple thread.(Patrick Zhai)
Improvements
- Create a TaskExecutor when an executor is not provided to the IndexSearcher, in order in order to simplify consumer's code (Luca Cavanna)code
- Speed up Panama vector support, use FMA, and test improvements. (Uwe Schindler, Robert Muir)
- FSTCompiler can now approximately limit how much RAM it uses to share
suffixes share suffixes during FST construction using the suffixRAMLimitMB method. (Mike McCandless)
Optimizations
- Faster top-level conjunctions on term queries when sorting by descending by descending score. (Adrien Grand)
- Change Postings back to using FOR in Lucene99PostingsFormat. Freqs, positions and offset keep using PFOR. (Jakub Slowinski)
Further details of changes are available in the change log available at: http://lucene.apache.org/core/9_9_0/changes/Changes.html
...